Apple quietly disclosed a curious deal in a regulatory filing submitted to the European Commission: it has entered into an agreement with the personalized audio startup Huxe Huxe to hire some of its employees andnon-exclusively licenseits intellectual property. This “hire talent without acquiring the company” model has a specific name—reverse acqui-hire. Huxe’s background points directly to a rapidly heating sector:AI-generated personalized podcasts.
Who is Huxe? A pioneer in “AI podcasts”
Huxe’s founding team is no ordinary group—they previously developed NotebookLMGoogle Gemini’s Gemini Notebook(now rebranded as AI-generated podcastfeatures at Google.NotebookLM Google Gemini’s “Audio Overview” feature once went viral, automatically converting any document into a dialogue-style show hosted by two AI presenters—many people’s first realization that “AI podcasts” are not science fiction.
After leaving Google, this team founded Huxe to focus on more advanced “personalized audio” experiences. But the company did not survive: on May 21, Huxe announcedclosureits shutdown, removing its app from the App Store and Google Play, ceasing all services, and deleting user data. In its farewell statement, the team wrote: “The team is moving on to new things—we will not continue developing this product.”
Apple’s “reverse acqui-hire” strategy
Less than three weeks after Huxe’s closure, Apple notified the European Commission of this deal on June 9. The so-called “reverse acqui-hire” is a tactic increasingly adopted by major companies:hiring only key team members and licensing the technology—without acquiring the startup as a wholeThis allows Apple to quickly acquire AI talent and technology while avoiding antitrust scrutiny that a full acquisition might trigger.
Regulatory filings do not disclose which individuals received offers, whether those offers were accepted, or how Apple plans to deploy Huxe’s technology. Yet the timeline makes Apple’s intent quite clear.
Intent revealed by the timeline:对标 Spotify
One detail is worth noting: the day before Huxe announced its shutdown,Spotify had just launched its own AI podcast generation feature.In other words, in the same week Spotify made a high-profile entry into AI-powered podcasts, Apple targeted precisely the team “most skilled at building AI podcasts.”
Industry insiders widely speculate Apple wants to embed similar functionality into its Podcasts app— enabling users to generate personalized AI audio programs based on their subscriptions, news feeds, and documents. Apple executives appear deeply fascinated by the concept of “personalized audio as a replacement for traditional podcasts.”While it remains unclear whether mainstream listeners will embrace this idea, the viral spread of NotebookLM audio summaries demonstrates genuine user demand for this direction.
Reverse acquisition: the new normal for AI talent hoarding
Another notable aspect of Apple’s deal is that “reverse acquisition” is becoming a standard tactic for AI giants to accumulate talent. Over the past year, Microsoft, Google, and Meta have all used similar approaches—selectively poaching core teams from competitors or startups while sidestepping the regulatory scrutiny and lengthy integration processes associated with full acquisitions. For AI startups struggling to secure substantial funding or achieve independent sustainability, being “reverse-acquired” by a tech giant has increasingly become a dignified exit path—teams transition en masse, while products are quietly wound down.
In this AI arms race, Apple has long been criticized externally for “lagging behind”—lacking its own large language model and rarely discussing its generative AI strategy publicly. Yet through precise reverse acquisitions like the one with Huxe, Apple is steadily filling capability gaps in vertical domains such as AI audio and AI assistants via a “buy talent + buy technology” approach. This low-key, pragmatic method aligns more closely with Apple’s longstanding ethos than flashy concept-product launches.
AI audio is emerging as a new battleground
Apple’s deal thrusts “AI audio”—a previously under-the-radar niche—into the spotlight. Looking back over the past year, this space has already grown remarkably active: Google gained traction with NotebookLM audio summaries, Spotify directly integrated AI podcast generation, and Apple now addresses its talent gap through reverse acquisition. The moves by these three tech giants in AI audio mirror their competition inlarge language models— except this time, the battleground is “what you listen to during your daily commute.”
For average users, the most likely change is:In the near future, your podcast app will feature a “Generate My Podcast” buttonEnter a few topics you’re interested in, and AI will create a custom show exclusively for you. This is already technically feasible; what’s missing is product refinement and copyright compliance clarification.
Frequently Asked Questions (FAQ)
What is Huxe?
Huxe is a startup building “personalized AI audio,” whose founding team previously worked at Google on NotebookLM AI-generated podcast features. The company shut down in May 2026; Apple subsequently hired some of its employees and licensed its technology.
How does a reverse acquisition differ from a standard acquisition?
A standard acquisition involves purchasing the entire company; a reverse acqui-hire involves hiring only key team members and licensing the technology—without acquiring the company itself. This approach helps avoid antitrust scrutiny and eliminates the cost and complexity of integrating the entire organization.
What does Apple intend to do with the Huxe team?
Apple has not publicly disclosed its plans, but industry consensus speculates it aims to integrate AI-generated podcast capabilities into its Podcasts app—competing directly with Spotify’s similar offering. No official announcement of specific plans has been made.
How do AI-generated podcasts differ from traditional podcasts?
Traditional podcasts are recorded by humans; AI podcasts are automatically generated by models based on documents, topics, or news you provide—the host’s voice is synthesized, content is fully personalized, and episodes can be updated instantly and generated infinitely.
Want to Discover More Useful AI Tools? Explore Our AI Model Library and Tool Comparison Engineor continue reading:Google gives Gemini a face · Gemini 4 Argon released · Kuaishou Kling in-depth review.